Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
What it really takes to learn how to code? (sailsoftware.co)
13 points by alexp on Sept 9, 2015 | hide | past | favorite | 2 comments


The best way to learn how to code is have an idea of what you want to build and then seek out the knowledge you need to build it. I personally never learned from any tutorial until I wanted actually to build something. Keep working until you figure out how to solve each problem at each step. You'll get better and faster over time.


This is a little bit of an aside, but it's something I've been thinking about lately:

Only professional programmers program because we design our libraries and tools for a professional "workflow". We've all scheduled a few weeks per year for tooling, so every tool presumes you can spend some time integrating it into your tooling, and that you have a general understanding of how to do that. And everything assumes you have a bunch of ancillary tooling already set up. React presumes you already have an editor and a web host. Heroku assumes you already have a framework set up. There's nowhere you can just step in and casually play with useful, running software.

In other words, everything is designed with the presumption that you are a full time programmer. We have built a high wall around us in the form of tooling that takes 6mo to a year of full time effort to start to be able to solve generic problems.

As a side effect, the complicated tooling makes us all feel powerful and special... most of us quite like feeling like a wizard! And by keeping everything "in the profession" we keep our wages high. I think we might be a little let down if programming was demystified for everyone else. Most of us push towards the most complex problems we can manage, rather than making basic stuff doable by people who are just casually interested in programming.

But it's a chicken and egg problem. There are no casual programmers so there is no market you can easily sell into if you did. No market = no tools. No tools = no market.

I think it will change within the next decade (I've been trying for the last decade). Several people are making attempts. But because it's an ecosystem problem it's really challenging to get your MVP scope right.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: